Ready, Set, Runtime

AI lets attackers and developers move faster than any human team can review. So who's responsible for security when agents write the code? In Episode 4 of Ready, Set, Runtime (Season 1: The New SDLC), AHEAD Field CTO Michael "MK" Kolbrener talks with Tim Curless, Principal Architect for AI Engineering at AHEAD, and Felix Vargas, Senior Director of Security Specialist Solution Engineering at AHEAD, about what security looks like when the pipeline moves faster than the review cycle.

"I think we need to shift down so that it's covered in the platform, and that is then applied the same way to agents or to humans." – Tim Curless, Principal Architect for AI Engineering, AHEAD

AI is speeding up both sides of the fight. Attackers no longer need deep expertise to run sophisticated attacks, and development teams are producing more code than people can realistically review. Felix and Tim agree the answer to "who's responsible?" hasn't changed: security is everyone's job, with the CISO managing the risk. What has changed is speed. Felix makes the case for running the fundamentals at machine speed, from discovery and exploitability validation to agents in the SOC. Tim explains why the controls need to live in the platform itself, so the same guardrails apply to every developer and every agent.

The conversation also covers:

• Identity for non-human actors: accountable owners, scoped tokens, immutable audit records, and behavioral analytics
• The pull request bottleneck, and an iceberg model where agents handle routine review and humans own business logic
• Secrets in code, software supply chain risk, and taint propagation between agents
• Prioritizing exposures by exploitability and business risk instead of chasing every finding at once
• Why citizen developers need security built into the platform they build on

They close with what clients should focus on in the next 12 months, and why a deterministic governance layer has to ride underneath the probabilistic layer of AI.

About the Guests Tim Curless is Principal Architect for AI Engineering at AHEAD. Tim leads enterprise AI platform engineering, bringing AI into the software development lifecycle to improve security, governance, and effectiveness, and builds agentic systems such as DevSecOps Sentinel. Tim is also building a quantum computing lab on NVIDIA DGX. Felix Vargas is Senior Director of Security Specialist Solution Engineering at AHEAD. With 23 years in technology, Felix leads AHEAD's security pre-sales team, advising enterprises on securing their AI platforms and using AI to strengthen cyber defense.

About the Host Michael "MK" Kolbrener is Field CTO at AHEAD, focused on Financial Services and Capital Markets, with more than 20 years of experience delivering enterprise technology, and the host of Ready, Set, Runtime.

About the Series Ready, Set, Runtime is an AHEAD podcast featuring technology leaders who are actively shaping the industry. Each episode explores emerging innovations across key technology domains and their impact on enterprise practices, strategies, and operations, as seen through the eyes of those doing the work. Season 1: The New SDLC is an eight-episode discussion following AI's impact across the full software development lifecycle, from accelerated development and agentic engineering to security, economics, enterprise deployment, and infrastructure. No vendor pitch. No hype. Just practitioner-level conversation.

Have a question about your own environment? Talk to AHEAD: https://www.ahead.com

An AI agent can turn out a pull request in 15 minutes. So why does it still wait three days for review? In Episode 3 of Ready, Set, Runtime (Season 1: The New SDLC), AHEAD Field CTO Priyanka Karan takes the host chair to explore where the developer experience ends and enterprise delivery begins. Joining the conversation are David Pan, Field CTO at SpaceX Super Intelligence (SpaceXSI), who came over in the Cursor acquisition; Martin Reynolds, Field CTO at Harness; and series host Michael "MK" Kolbrener.

"The ones that are really making a difference are where they don't treat AI as a tool. They treat AI like a delivery system change." – Martin Reynolds, Field CTO, Harness

AI coding tools have made writing code dramatically faster, but writing code was never most of the job. David and Martin bring two vantage points: the tool closest to the individual developer, and the platform closest to enterprise delivery. Together they explain why individual gains don't automatically add up to organizational ones. David shares Cursor platform data showing the top 1% of engineers pushing roughly 15 times as many pull requests as the median engineer. Martin explains why the organizations pulling ahead are the ones automating everything downstream of the pull request.

The conversation also covers:

• How teams are shrinking from two-pizza teams to two- and three-person, agent-native pods
• Why specialists may shift to building the shared skills and agents that generalists rely on
• How guardrails can work as a permission structure instead of a roadblock
• Why AI may end up doing more defense than offense, from testing and security reviews to automated 12-hour vulnerability remediation

They close on Agile fatigue, the debate over small versus big pull requests, how to train junior engineers who have never fixed bugs by hand, and an honest three-year forecast for enterprise adoption.

About the Guests David Pan is a Field CTO at SpaceX Super Intelligence (SpaceXSI), having joined through the Cursor acquisition, with a career spent in software engineering and software engineering leadership. Martin Reynolds is a Field CTO at Harness. Before joining Harness about three years ago, Martin spent 30 years delivering software across a wide range of organizations, and was one of Harness's first customers worldwide and its first in Europe. Michael "MK" Kolbrener is Field CTO at AHEAD, focused on Financial Services and Capital Markets, with more than 20 years of experience delivering enterprise technology, and the host of Ready, Set, Runtime.

About the Host Priyanka Karan is a Field CTO at AHEAD with two decades of experience in software development and digital transformation, focused on what those shifts mean for developers and AHEAD's clients.

About the Series Ready, Set, Runtime is an AHEAD podcast featuring technology leaders who are actively shaping the industry. Each episode explores emerging innovations across key technology domains and their impact on enterprise practices, strategies, and operations, as seen through the eyes of those doing the work. Season 1: The New SDLC is an eight-episode discussion following AI's impact across the full software development lifecycle, from accelerated development and agentic engineering to security, economics, enterprise deployment, and infrastructure. No vendor pitch. No hype. Just practitioner-level conversation.

Have a question about your own environment? Talk to AHEAD: https://www.ahead.com

A convincing AI demo is easier to build than ever. So what makes enterprise AI deployment still so hard? In Episode 2 of Ready, Set, Runtime (Season 1: The New SDLC), AHEAD Field CTO Michael "MK" Kolbrener talks with Kurt Boden, Principal Technical Consultant, and Habeeba Mansour, Senior Technical Consultant, about the gap between the demo and the deploy.

"We went from 'I need a month to be able to create something that I can show to other people' to 'if you just give me 24 hours, you can turn something around that looks pretty good.'" – Habeeba Mansour, Senior Technical Consultant

Two practitioners who build AI solutions for enterprise clients explain how fast the work has changed. A working prototype can now come together in a day, and the bar for a good demo keeps rising. That speed brings the harder questions sooner: who governs the agents, who owns them, and who keeps watch once they are live.

Along the way, Kurt and Habeeba share how their own use of AI has evolved over 18 months, from approving every step to trusting it with heavier lifts; why they evaluate tools by building with them; and how differently they think about token usage. They also dig into agent sprawl, ownership and identity for AI agents, whether MCP will last, and what comes next, with an optimist and a more cautious voice in the same conversation.

About the Host

Michael "MK" Kolbrener is Field CTO at AHEAD, focused on Financial Services and Capital Markets, with more than 20 years of experience delivering enterprise technology.

About the Series

Ready, Set, Runtime is an AHEAD podcast featuring technology leaders who are actively shaping the industry. Each episode explores emerging innovations across key technology domains and their impact on enterprise practices, strategies, and operations, as seen through the eyes of those doing the work.

Season 1: The New SDLC is an eight-episode discussion following AI's impact across the full software development lifecycle, from accelerated development and agentic engineering to security, economics, enterprise deployment, and infrastructure. No vendor pitch. No hype. Just practitioner-level conversation.

Have a question about your own environment? Talk to AHEAD: https://www.ahead.com

Is AI-accelerated development just faster code, or a whole new way of delivering technology across the enterprise? In the debut episode of Ready, Set, Runtime (Season 1: The New SDLC), AHEAD Field CTO Michael "MK" Kolbrener sits down with Josh Perkins and Craig Martin to break down what has actually changed.

"18 months at the current rate of change is equivalent to five years a few years ago." – Josh Perkins, VP, Emerging Technologies

AI is reshaping not just what software does, but how it gets planned, built, secured, and paid for. MK, Josh, and Craig trace accelerated development from code generation to a digital workforce of AI agents, and explain why the software development lifecycle (SDLC) is the starting point for something much bigger across the business.

They get candid about the worst practices they've seen: flooding teams with low-quality generated code, mistaking code generation for software engineering, and waiting on the sidelines out of fear of getting it wrong. They also cover where to start (business outcomes, not tools), why technical debt is one of the smartest first targets, how DevOps maturity and data readiness predict success, and what comes next: context engineering, tokenomics, and building enough trust to hand real work to agentic AI.

Chapters

00:00 Intro: MK, Josh Perkins & Craig Martin

01:52 What Is AI-Accelerated Development?

07:45 From SDLC to XDLC: Beyond Software

09:12 What's Driving the Shift: Speed, Tech Debt & Adoption

14:54 The Worst Practices in Adopting AI Dev Tools

21:47 Where to Start: Business Outcomes Over Tools

31:47 The Next 18 Months: Context Engineering, Tokenomics & Trust

41:41 What's Most Exciting in Emerging Tech Right Now

About the Guests

Josh Perkins is Vice President of Emerging Technologies at AHEAD, where he helps shape the company's AI strategy and works with enterprise leaders to turn emerging technologies into scalable capabilities. His work spans AI platforms, infrastructure, and operating models. He previously served as a Field CTO at AHEAD.

Craig Martin is Vice President of Advanced Engineering at AHEAD, where he helps clients turn strategy into execution through the architecture, engineering models, and delivery patterns needed to modernize applications, platforms, and operations. Before AHEAD, he led large-scale cloud modernization efforts and founded an AI-focused startup built to help athletes reach their goals.

About the Host

Michael "MK" Kolbrener is Field CTO at AHEAD, focused on Financial Services and Capital Markets. He has spent more than 20 years delivering enterprise technology and works at the intersection of technology and business strategy.

About the Series

Ready, Set, Runtime is an AHEAD podcast featuring technology leaders who are actively shaping the industry. Each episode explores emerging innovations across key technology domains and their impact on enterprise practices, strategies, and operations, as seen through the eyes of those doing the work.

Season 1: The New SDLC is an eight-episode discussion following AI's impact across the full software development lifecycle, from accelerated development and agentic engineering to security, economics, enterprise deployment, and infrastructure. No vendor pitch. No hype. Just practitioner-level conversation.

Have a question about your own environment? Talk to AHEAD: https://www.ahead.com

Ready, Set, Runtime is a podcast from AHEAD about how AI is changing the way software gets built, secured, shipped, and paid for.

The questions go well beyond code generation. What happens to Agile when delivery speeds up? How do teams govern AI without slowing down? What does it take to move from experimentation to deployment at scale?

Host Michael "MK" Kolbrener, AHEAD's Field CTO for Financial Services and Capital Markets, brings those questions to the practitioners and industry experts doing the work. Topics this season include:

• Accelerated development and agentic engineering
• Accountability and security when AI writes the code
• What AI actually costs, and how leaders decide what it's worth
• Closing the gap between the demo and enterprise deployment
• The developer tooling and infrastructure underneath it all

No vendor pitch. No hype. Follow wherever you listen so you don't miss an episode.

About AHEAD

AHEAD builds and manages digital platforms that power the most successful organization in the world. Our consultative approach, unmatched engineering, and innovative solutions combine to accelerate the impact of technology in every client we serve. By stitching together advances in Cloud, Automation, Operations, Security, FinOps and DevOps, we help clients deliver on the promise of digital transformation.

Learn more by visiting us at www.ahead.com